{"id":137136,"date":"2021-06-21T13:55:32","date_gmt":"2021-06-21T10:55:32","guid":{"rendered":"https:\/\/www.enerjigazetesi.ist\/?p=137136"},"modified":"2021-06-21T13:55:32","modified_gmt":"2021-06-21T10:55:32","slug":"bilgi-islemeye-dayali-gorusme-ozetleyen-yapay-zeka-gelistirildi","status":"publish","type":"post","link":"https:\/\/www.enerjigazetesi.ist\/en\/bilgi-islemeye-dayali-gorusme-ozetleyen-yapay-zeka-gelistirildi\/","title":{"rendered":"(Turkish) Bilgi \u0130\u015flemeye Dayal\u0131 G\u00f6r\u00fc\u015fme \u00d6zetleyen Yapay Zeka Geli\u015ftirildi"},"content":{"rendered":"

Sorry, this entry is only available in Turkish<\/a>. For the sake of viewer convenience, the content is shown below in the alternative language. You may click the link to switch the active language.<\/p>

Mitsubishi Electric Corporation, s\u00f6zl\u00fc g\u00f6r\u00fc\u015fmeleri isabetli bir \u015fekilde otomatik olarak \u00f6zetleyen bir teknoloji geli\u015ftirdi. Yapay zeka markas\u0131 MAISART\u00ae* temelli bir g\u00f6r\u00fc\u015fme \u00f6zetleme teknolojisi olan bu sistem, Mitsubishi Electric\u2019in \u00e7a\u011fr\u0131 merkezinde yap\u0131lan \u00f6n testlere g\u00f6re \u00e7al\u0131\u015fanlar\u0131n \u00e7a\u011fr\u0131 raporu haz\u0131rlamalar\u0131 i\u00e7in harcad\u0131klar\u0131 s\u00fcreyi yakla\u015f\u0131k olarak yar\u0131 yar\u0131ya azalt\u0131yor.<\/strong><\/p>\n

*<\/span>(M<\/u>itsubishi Electric’s AI<\/u> creates the S<\/u>tate-of-the-ART<\/u> in technology)<\/span><\/em><\/p>\n

\"\"

\u015eekil 1: \u00c7a\u011fr\u0131 merkezinde uygulama \u00f6rne\u011fi<\/p><\/div>\n

\u00d6ne \u00c7\u0131kan \u00d6zellikler<\/strong><\/p>\n

1) Yaz\u0131l\u0131 kay\u0131tlar\u0131 otomatik olarak olu\u015fturuyor ve rapor haz\u0131rlama s\u00fcresini yar\u0131 yar\u0131ya d\u00fc\u015f\u00fcr\u00fcyor<\/strong><\/h2>\n

\u00c7a\u011fr\u0131 merkezi g\u00f6r\u00fc\u015fmeleri<\/strong>, genellikle geleneksel konu\u015fma tan\u0131ma teknolojisiyle kay\u0131t alt\u0131na al\u0131n\u0131yor. Bu kay\u0131tlar, \u00e7o\u011funlukla anla\u015f\u0131lmayan s\u00f6zler<\/strong> veya kopuk konu\u015fmalar<\/strong> i\u00e7erdi\u011finden bir yaz\u0131l\u0131m taraf\u0131ndan tam isabetli olarak yaz\u0131l\u0131 kay\u0131tlara d\u00f6n\u00fc\u015ft\u00fcr\u00fclemiyor. Mitsubishi Electric’in yeni teknolojisi (\u015eekil 2), g\u00fcnl\u00fck konu\u015fma dili<\/strong> ve e\u015f anlaml\u0131 ifadeler<\/strong> de dahil olmak \u00fczere, anlamlar\u0131 do\u011fru bir \u015fekilde belirlemek i\u00e7in diyalog ba\u011flamlar\u0131n\u0131<\/strong> \u00f6\u011freniyor. Ge\u00e7mi\u015f raporlarda yer alan verileri kullanarak kopuk konu\u015fmalar, dil bilgisi hatalar\u0131<\/strong> ve kelime tercihlerini<\/strong> topluyor, tam ve do\u011fal c\u00fcmleler olu\u015fturuyor. Daha sonra \u00e7al\u0131\u015fanlar\u0131n \u00fczerinde d\u00fczenlemeler yapt\u0131\u011f\u0131 ge\u00e7mi\u015f raporlardan en benzer ve en k\u0131sa c\u00fcmleleri \u00e7\u0131kar\u0131yor ve en sonunda \u00f6zet yaz\u0131l\u0131 kayd\u0131 tamaml\u0131yor. Yap\u0131lan \u015firket i\u00e7i testler, Mitsubishi Electric’in yeni teknolojisinin Japonca metin hacmini<\/strong> b\u00fcy\u00fck oranda azalt\u0131rken metnin yakla\u015f\u0131k % 90<\/strong>‘\u0131n\u0131n (geleneksel teknolojide yakla\u015f\u0131k y\u00fczde 30) kullan\u0131labilir hale getirildi\u011fini g\u00f6steriyor. Bu sayede, \u00e7a\u011fr\u0131 merkezi \u00e7al\u0131\u015fanlar\u0131n\u0131n raporlar\u0131n\u0131<\/strong> manuel haz\u0131rlamak i\u00e7in harcad\u0131klar\u0131 s\u00fcre yar\u0131 yar\u0131ya<\/strong> azal\u0131yor.<\/p>\n

2) Zaman i\u00e7inde d\u00fczeltme isabetlili\u011fini iyile\u015ftirmek i\u00e7in manuel d\u00fczeltmeleri \u00f6\u011freniyor<\/strong><\/h2>\n

\u0130nsanlar yeni \u00fcr\u00fcnler hakk\u0131nda bilgi almak i\u00e7in bir \u00e7a\u011fr\u0131 merkezini arad\u0131klar\u0131nda<\/strong> geleneksel yaz\u0131l\u0131mlar benzeri konular\u0131 kapsayan \u00e7a\u011fr\u0131 ge\u00e7mi\u015flerine<\/strong> ve ge\u00e7mi\u015f raporlara<\/strong> ba\u015fvuram\u0131yorlar. Bu nedenle g\u00f6r\u00fc\u015fmelerin isabetli kay\u0131tlar\u0131n\u0131n otomatik<\/strong> olarak olu\u015fturulmas\u0131 m\u00fcmk\u00fcn<\/strong> olmuyor. Mitsubishi Electric’in yeni yapay zek\u00e2 teknolojisi<\/strong>; \u00e7al\u0131\u015fanlar\u0131n kopuk s\u00f6z dizimleri<\/strong>, g\u00fcndelik jargonu<\/strong> veya e\u015f sesli s\u00f6zc\u00fckleri<\/strong> daha do\u011fal bir dile manuel olarak nas\u0131l d\u00f6n\u00fc\u015ft\u00fcrd\u00fcklerini \u00f6\u011frenmek i\u00e7in son d\u00fczenlenmi\u015f raporlara ba\u015fvuruyor<\/strong> ve bilgisini s\u00fcrekli olarak g\u00fcncelliyor.<\/p>\n

\"\"

\u015eekil 2: \u00d6zetleme s\u00fcreci<\/p><\/div>\n

Yeni ve geleneksel y\u00f6ntemlerin kar\u015f\u0131la\u015ft\u0131rmas\u0131<\/strong><\/h2>\n\n\n\n\n\n
 <\/td>\n\n

G\u00f6r\u00fc\u015fme i\u015fleme y\u00f6ntemi\u00a0<\/strong><\/p>\n<\/td>\n

\n

Performans<\/strong><\/p>\n<\/td>\n<\/tr>\n

\n

Yeni<\/strong><\/p>\n<\/td>\n

MAISART, yaz\u0131l\u0131 kay\u0131t olu\u015fturmak amac\u0131yla \u00e7a\u011fr\u0131 ge\u00e7mi\u015flerinden ve ge\u00e7mi\u015f raporlardan \u00f6\u011freniyor ve manuel haz\u0131rlanm\u0131\u015f ge\u00e7mi\u015f raporlardan en benzer ve en k\u0131sa c\u00fcmleleri bularak \u00f6zetlenmi\u015f do\u011fal c\u00fcmlelerden k\u0131sa kay\u0131tlar olu\u015fturuyor.<\/td>\nOlu\u015fturulan kayd\u0131n yakla\u015f\u0131k y\u00fczde 90’\u0131 son raporlarda kullan\u0131labiliyor.<\/td>\n<\/tr>\n
Geleneksel<\/strong><\/td>\nYapay zek\u00e2, u\u00e7tan uca makine \u00f6\u011frenmesiyle s\u00f6zl\u00fc c\u00fcmlelerin \u00f6zet c\u00fcmlelere nas\u0131l d\u00f6n\u00fc\u015ft\u00fcr\u00fclece\u011fini \u00f6\u011freniyor.<\/td>\nOlu\u015fturulan kayd\u0131n yakla\u015f\u0131k y\u00fczde 30’u son raporlarda kullan\u0131labiliyor.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Gelecek i\u00e7in planlar<\/strong><\/h2>\n

Mitsubishi Electric, bu yeni teknolojisini Mart 2021\u2019den itibaren<\/strong> kendi \u00e7a\u011fr\u0131 merkezinde demo testleriyle de\u011ferlendirmeye devam ediyor. Bu test sonu\u00e7lar\u0131n\u0131 \u00fcr\u00fcn hatalar\u0131<\/strong> ve \u00fcr\u00fcnler<\/strong> hakk\u0131nda m\u00fc\u015fterilerle yap\u0131lan telefon g\u00f6r\u00fc\u015fmelerinin yaz\u0131l\u0131 kay\u0131tlar\u0131n\u0131<\/strong> olu\u015fturmada kullanmay\u0131 ama\u00e7l\u0131yor.<\/p>\n

Arka plan<\/strong><\/h2>\n

Japon h\u00fck\u00fcmeti<\/strong> taraf\u0131ndan 2019 y\u0131l\u0131nda \u00e7a\u011fr\u0131 merkezlerine<\/strong> ili\u015fkin yay\u0131nlanan bir ara\u015ft\u0131rmaya g\u00f6re, ankete kat\u0131lanlar\u0131n % 41,8<\/strong>‘i operat\u00f6r\u00fcn \u00e7al\u0131\u015fma verimlili\u011finin art\u0131r\u0131lmas\u0131n\u0131n<\/strong> en \u00f6nemli konu oldu\u011funu ifade etti. Bu nedenle, m\u00fc\u015fterilerle yap\u0131lan telefon g\u00f6r\u00fc\u015fmelerine<\/strong> ili\u015fkin raporlar\u0131n haz\u0131rlanma s\u00fcrecinde otomasyona y\u00f6nelik<\/strong> talep art\u0131yor. Geleneksel teknolojilerle y\u00fcr\u00fct\u00fclen m\u00fc\u015fteri g\u00f6r\u00fc\u015fmelerine ili\u015fkin bilgilerin son raporlarda kullan\u0131labilirli\u011fi ancak % 30<\/strong> d\u00fczeylerinde kal\u0131yor. Sonu\u00e7 olarak, bu raporlar\u0131n geleneksel yapay zek\u00e2 yerine manuel<\/strong> olarak haz\u0131rlanmas\u0131 operat\u00f6rler i\u00e7in daha pratik bir uygulama<\/strong> olmaya devam ediyor.<\/p>","protected":false},"excerpt":{"rendered":"

Sorry, this entry is only available in Turkish. For the sake of viewer convenience, the content is shown below in the alternative language. You may click the link to switch the active language.Mitsubishi Electric Corporation, s\u00f6zl\u00fc g\u00f6r\u00fc\u015fmeleri isabetli bir \u015fekilde otomatik olarak \u00f6zetleyen bir teknoloji geli\u015ftirdi. Yapay zeka markas\u0131 MAISART\u00ae* temelli bir g\u00f6r\u00fc\u015fme \u00f6zetleme teknolojisi […]<\/p>\n","protected":false},"author":1,"featured_media":137140,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[51,53],"tags":[53114,52259,46449,81172,79170,65918,61765],"views":140,"_links":{"self":[{"href":"https:\/\/www.enerjigazetesi.ist\/en\/wp-json\/wp\/v2\/posts\/137136"}],"collection":[{"href":"https:\/\/www.enerjigazetesi.ist\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.enerjigazetesi.ist\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.enerjigazetesi.ist\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.enerjigazetesi.ist\/en\/wp-json\/wp\/v2\/comments?post=137136"}],"version-history":[{"count":0,"href":"https:\/\/www.enerjigazetesi.ist\/en\/wp-json\/wp\/v2\/posts\/137136\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.enerjigazetesi.ist\/en\/wp-json\/wp\/v2\/media\/137140"}],"wp:attachment":[{"href":"https:\/\/www.enerjigazetesi.ist\/en\/wp-json\/wp\/v2\/media?parent=137136"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.enerjigazetesi.ist\/en\/wp-json\/wp\/v2\/categories?post=137136"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.enerjigazetesi.ist\/en\/wp-json\/wp\/v2\/tags?post=137136"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}